DocumentCode
3188613
Title
Test scheduling for modular SOCs in an abort-on-fail environment
Author
Ingelsso, Urban ; Goel, Sandeep Kumar ; Larsson, Erik ; Marinissen, Erik Jan
Author_Institution
Dept. of Comput. Sci., Linkopings Univ., Linkoping, Sweden
fYear
2005
fDate
22-25 May 2005
Firstpage
8
Lastpage
13
Abstract
Complex SOCs are increasingly tested in a modular fashion, which enables us to record the yield-per-module. In this paper, we consider the yield-per-module as the pass probability of the module´s manufacturing test. We use it to exploit the abort-on-fail feature of ATEs, in order to reduce the expected test application time. We present a model for expected test application time, which obtains increasing accuracy due to decreasing granularity of the abortable test unit. For a given SOC, with a modular test architecture consisting of wrappers and disjunct TAMs, and for given pass probabilities per module test, we schedule the tests on each TAM such that the expected test application time is minimized. We describe two heuristic scheduling approaches, one without and one with preemption. Experimental results for the ITC´02 SOC test benchmarks demonstrate the effectiveness of our approach, as we achieve up to 97% reduction of the expected test application time, without any modification of the SOC or ATE.
Keywords
automatic test equipment; integrated circuit testing; integrated circuit yield; scheduling; system-on-chip; abort-on-fail environment; automatic test equipment; heuristic scheduling; integrated circuit testing; manufacturing test; modular SOC; system-on-chip testing; test access mechanism; test scheduling; yield-per-module; Application specific integrated circuits; Benchmark testing; Circuit testing; Costs; Integrated circuit testing; Job shop scheduling; Pulp manufacturing; Semiconductor device manufacture; Semiconductor device testing; System testing;
fLanguage
English
Publisher
ieee
Conference_Titel
Test Symposium, 2005. European
Print_ISBN
0-7695-2341-2
Type
conf
DOI
10.1109/ETS.2005.38
Filename
1430002
Link To Document